A phase 1 trial of umbilical cord blood-derived tumor-reactive PD-L1+natural killer cells engineered to express soluble IL-15 (TRACK-NK) in patients with non-small-cell lung cancer (NSCLC) refractory to PD-1/PD-L1 inhibitors

TPS2665 Background: The mechanism by which patients with PD-L1 – tumors can respond favorably to anti-PD-L1 monoclonal antibody therapy has been shown to involve the PD-L1+ NK cell in the tumor microenvironment (TME). PD-L1 is induced on the cell surface when NK cells recognize tumors, leading to enhanced NK-cell function without exhaustion. The PD-L1 + NK cell is capable of trafficking to the lung TME, recognizing and killing tumor cells as well as inducing activation of endogenous T cells via potent cytokine secretion, all without the expression of a chimeric antigen receptor (CAR). TRACK-NK cells are PD-L1+ NK cells derived from cord blood (CB) and engineered to express soluble IL-15 (sIL-15). The TRACK-NK cells express high levels of tumor reactive receptors that recognize "tumor stress" including DNAM-1, NKp30 and NKG2D. TRACK-NK cells induced significant reduction in tumor volume in mice injected with human A549 NSCLC cells compared to mock-transduced NK cells, or NK cells expressing sIL-15 (sIL-15-NK) but without ex vivo activation. Survival was also prolonged in the H460 NSCLC model compared to vehicle alone. We hypothesized that TRACK-NK cells can be safely administered, and will provide meaningful clinical benefit in patients with NSCLC. Methods: NCT05334329 is a first-in-human, phase 1 dose finding study of allogeneic, off-the shelf frozen and thawed TRACK-NK cells in patients with NSCLC progressing on treatment with PD-1/PD-L1 inhibitors. TRACK-NK cells (CB NK cells engineered with retroviral transduction to express sIL-15 and ex vivo expanded and activated to express PD-L1) are administered in four weekly intravenous infusions per cycle following lymphocyte depletion (Fludarabine/Cyclophosphamide x 3 days) in the outpatient setting. Patients not experiencing disease progression (per RECIST v1.1) following cycle 1 may receive a second cycle. Primary endpoints include: 1. Toxicity/adverse events as per NCI-CTCA version 5.0, and ASTCT Consensus Grading for Cytokine Release Syndrome (CRS) and Neurotoxicity; and 2. Magnitude/duration of TRACK-NK cell persistence in peripheral blood via droplet digital PCR. We are employing a Utility Based-Bayesian Optimal Interval design to direct optimal dose determination. Dose limiting toxicity is defined as any ≥ grade 3 non-hematologic toxicity; any grade 3 CRS that does not resolve to ≤ grade 1 within 7 days; any ≥ grade 4 CRS; any ≥ grade 2 neurotoxicity that does not resolve to grade 1 within 72 hours; any ≥ grade 3 autoimmune, hypersensitivity, and acute graft-versus-host reactions. Correlatives also include cytokines levels, quantity and activation of CD4 and CD8 cells in peripheral blood, and NK and T cell trafficking and TME changes in paired tumor biopsies using quantitative immunofluorescence and gene expression analysis. Clinical trial information: NCT05334329 .